Cerebral blood flow (CBF) defines the rate of delivery of metabolic substrates and is central to brain metabolism. A noninvasive method for measuring regional cerebral blood flow is therefore important for assessing pathologies associated with alterations in cerebral metabolism such as carotid artery stenosis, stroke and brain tumors. The applicants proposed to develop a MR method for regional CBF and perfusion territory mapping and assess this technique in normal subjects and patients with carotid artery stenosis. The novel ASL technique for cerebral perfusion and perfusion territory mapping proposed in this exploratory development application will build on their preliminary MR continuous Arterial Spin Labeling (cASL) work which has provided the first system capable of performing cASL in humans at 3T. An array of three separately gated CW RF sources and independently detunable spin-labeling coils will be developed for placement over the left and right carotid and both vertebral arteries. This will improve MR perfusion measurements in four ways: (1) provide a magnetization transfer-free technique allowing the acquisition of multi-slice, arbitrary plane, quantitative MR perfusion maps; (2) improve sensitivity and coverage by covering each of the major arteries with a small labeling coil, thus enabling a multi-coil cASL method that will label both posterior and anterior circulation at 3T. Continuous ASL methods have previously been limited to 1.5T by potentially unsafe RF power demands; (3) independently labeling each of the major arteries to allow the mapping of the perfusion territory of each labeled artery. Thus, in addition to measuring the total level of blood flow to a region at risk for ischemic damage, we will develop a method capable of determining the regional flow fraction delivered by each of the three major sets of arterial inputs to the brain. The applicants will study the variability of the carotid and vertebral arteries' perfusion territories and overlap in watershed zones in normal subjects. In addition, they will study these parameters in patients with moderate (50-70%) unilateral carotid stenosis. This will allow them to determine the percentage of blood flow to a given cortical voxel from each of the three major arterial supplies providing a measure of collateral flow. This novel information could be valuable for guiding and following response to treatment of carotid artery stenosis.